A popular diary on the rec list purports to show how little an effect some proposed changes to how Social Security COLA's might get calculated would have on recipients. However, there is a fundamental error in the calculation that results in an underestimate of the effect by a factor of about 30 times!
The diarist makes a COLA calculation based on an example of a Social Security payment of $1034 per month and calculates a COLA of $34.35 per month based on a CPI change of 3.3%. Because the chained CPI would likely be around 0.4% less than the CPI that would make its change around 2.9%. So far so good. Now here's where the big error occurs. The diarist applies the 0.4% difference to the monthly COLA itself i.e. the $34.45 number to arrive at a $.14 difference. But the 0.4% difference should have been applied to the entire monthly payment of $1034. If done that way, that results in a difference per month of not $.14 but actually $4.14. That's a factor of 30x difference. And it compounds that way year after year every time there's a COLA calculation.
Another way of looking at it is as follows: if the chained CPI is 0.4% lower than the CPI when the CPI is 3.3% then the chained CPI results in a COLA reduction of 0.4%/3.3% x 100% or about 12%. That's a significant hit for seniors now and in the future.
Updated to correct my own math error! $.14 not $.014 LOL.